Management of IT-Enabled Business Projects Dr. Mary C. Lacity “Man, Universe, the Controller” Diego Rivera (1886-1957) 1 Management of IT-Enabled Business Projects The objectives of this module: Gain an understanding of the track record and challenges of managing large scale, IT-enabled projects Understand the relationship between: project management: managing the project phases change management: managing the people affected by change, mostly through training & incentives Explain robust, best practices for “doing the thing right”. These are management practices that have been associated with success since the beginning of IT—the practices do not change just because technologies change. Gain an appreciation for the difference between “doing the thing right” and 2 “Doing the right thing”. Session Objectives Business solutions drive technology selection Secure top management active support Select high powered project manager with proven Ensure to track record Value for IT Involve & incent knowledgeable users Spend Don't judge success solely based on time to budget Buy-in outside expertise to transfer learning Implement incrementally Manage scope creep, perhaps using 80/20 satisficing rule Beware of Mythical Man Month Doing The Thing Right: Robust Best Practices for Project Management 3 Relationship Between Project Management & Change Management PROJECT MANAGEMENT PHASES Planning Requirements Analysis Design Development Testing Implementation Hands-on Acceptance CHANGE PHASES Pre-awareness Awareness Self-concern Mental Tryout Change Strategy Audience Analysis & Impact Assessment Sponsorship & Communication Plan Role Design & Mapping Training Design Training Deployment 4 Readiness Assessment Source: Adapted from Roberts, Jarvenpaa & Baxley, MISQE, 2003 Startup Support Change Management The #1 contributor to project success is strong, visible and effective sponsorship. The top obstacle to successful change is employee resistance at all levels: front-line, middle managers, and senior managers. Employees want to hear messages about change from two people: the CEO or their immediate supervisor (and these messages are not the same). When asked what they would do differently next time, most teams would begin their change management activities earlier in their next project, instead of viewing it as an add-on or afterthought. The top reasons for employee resistance are a lack of awareness about the change, comfort with the ways things are and fear of the unknown. Middle managers resist change because of fear of losing control and overload of current tasks and responsibilities. Source: Report sponsored by ProSci based on 284 participants in 50 countries 5 Project Management Studies For the past 30 years, up to 80 % of all IT projects are late, over-budget, and/or fail to deliver promised functionality. See: Keil, Mark and Montealegre, Ramiro, “Cutting Your Losses,Extricating Your Organization When a Big Project Goes Awry,”Sloan Management Review, Spring , 2000, pp. 55-68. See: Sauer, Chris, Why Information Systems Fail, Alfred Waller,1993. See: “The Standish Group CHAOS Report” reported on: http://www.standishgroup.com/chaos.html See: Peter Morris, “Project Management: Lessons from IT and Non-IT Projects,” in Information Management: The Organizational Dimension, Earl (ed), Oxford University Press, 1996, pp. 321-336. 6 California Department of Social Services State-wide Automated Child Support System Expectations: $76 million contract with Lockheed Martin to develop the system 3 year schedule to complete Outcome: Cancelled in November 1997, after spending $435 million 7 Denver International Airport Baggage Handling System Expectations: $175.6 million contract with BAE Automated Systems to develop the system 18 month schedule to complete, April 22 1992 to Oct 1993 Outcome: Cancelled after horrible test in April 1994, after spending over $2 billion 8 London Stock Exchange Integrated Claims and Settlement System Expectations: £50 million cost estimate to develop the system (Cooper’s as project management) 2 year project to complete, Dec 1989 to Oct 1991 Progress: In-house staff will take too long, buy Vista software & hire Vista to modify (est. cost to modify, £1 million) Outcome: Cancelled on March 1993, after spending £400 million9 Standish Group CHAOS Report In the US, companies spend $250 billion each year on development of 175,000 IT projects. Research Method: Interviewed 365 IT executives about 8,380 IT projects. Large company: > $500 million in revenue Medium company: $200 to $500 million in revenue Small company: $100 to $200 million in revenue 10 Standish Group CHAOS Report: IT Projects classification Type I: Success; on time, on budget, promised functionality Type II: Challenged; over-budget, over-time and or missing functionality Type III:Failed; Severely impaired projects; cancelled projects 11 Standish Group CHAOS Report: IT Projects classification 12 Standish Group CHAOS Report: IT Projects classification SMALL MEDIUM LARGE TYPE I 28% 16.2% 9% TYPE II 50.4% 46.7% 61.5% TYPE III 21.6% 37.1% 29.5% 13 Standish Group CHAOS Report: Failure Statistics For every 100 project starts, there are 94 restarts For Type II and Type III Projects: delivered at 189% of original cost estimate: Large companies 178% Medium companies 183% Small companies 214% 14 Standish Group CHAOS Report: Failure Statistics For Type II and Type III Projects: delivered at 222% of original time estimate: Large companies 230% Medium companies 202% Small companies 239% 15 Standish Group CHAOS Report: Failure Statistics For Type II Projects: Only 61% of content delivered Large companies 42% Medium companies 65% Small companies 74% 16 Standish Group CHAOS Report: Failure Rates over Time 48 % of participants believe there are more failures today than 5 years ago. 46 % of participants believe that there are more failures today than 10 years ago 17 How do IT Project Success Rates Compare with other large projects? Peter Morris studied 1,444 projects world-wide in many industries. Only 2% of projects were delivered on budget Example: Trans Alaskan Pipeline estimated cost in 1969: $960 million actual cost: $8.7 billion He found that cost escalation was largely due to factors outside of project manager’s control, such as government action, inflation, strikes, technical uncertainty, scope changes, weather. 18 Morris Study Factors Affecting Project Success ENVIRONMENTAL FACTORS Politics, Community, Environment, Weather FINANCING & ECONOMICS poor c/b analysis budget cuts EXPECTATIONS/ ATTITUDES PROJECT DEFINITION Requirements known Technology Proven PROJECT OUTCOME IMPLEMENTATION best practices such as top management support, user involvement, adequate resources, effective teams, quality assurance 19 Industry Comparison of External Factors INDUSTRY EXTERNAL FACTORS AEROSPACE/DEFENSE POWER Stable requirements Long development (3 years); long life (30yrs) PETROCHEMICAL, MINING, SHIPPING As above but less protected by environmental factors CIVIL & BUILDING Criticized for poor management, but removed from users; products last years Intimate user involvement; multiple stakeholders; unstable requirements, changing specs soon as implemented IT PROJECTS 20 Standish Group CHAOS Report: Success Statistics on Type I DOING THE THING RIGHT: • User Involvement • Executive Management Support • Clear Statement of Requirements • Proper Planning • Realistic Expectations • Smaller Project Milestones • Competent Staff • Ownership 21 Doing the right thing…. Business drives, IT enables….. 22 Project Management: Three case studies Manufacturing: Centralized order scheduling system to reduce cycle time Centralized Decision-making Baking: Hardware migration to cut costs No changes in decision-making Insurance: Claims processing system to reduce fraud Decentralized decision making 23 Project Management: Three case studies Manufacturing: Expected: Costs of $6 million, 2 years Actual: $12 million, more than 4 years Baking: Expected: Costs of $7 million, 2 years Actual: $1.5 million for pilot, project cancelled Insurance: Expected: Costs of £4.5 million Actual: £6 million, almost a year late 24 Project Management: Three case studies Company U.S. Manufacturing U.S. Baking U.K. Insurance Sales in Company/SBU $1 billion Annual I.T. Budget $8 million # I.T. Employees 75 $2 billion $16 million 100 £650 million in £15 million insurance (about $24 premiums (appx. million) $1.1 billion) 180 25 QUESTIONS TO THINK ABOUT: Should success be based “on time” “on budget” “to functionality?” • Baking is a Type III project that met all objectives • Insurance is a Type II, mediocre results • Manufacturing is a Type II, excellent results Which projects “Did the thing right?” Which projects “Did the right thing?” 26 Old Claims Process at Insurance: Pay out £222,000,000 year Honey, what else can we add to this claim??? Throw in everything! Headquarters Mainframe 27 Old Claims Process at Insurance Headquarters Mainframe Claims typed in 28 Old Claims Process at Insurance Supervisor reads mainframe reports and assigns to a claims handler Headquarters Mainframe 29 Old Claims Process at Insurance Claims handler investigates the claim and decides on award Headquarters Mainframe 30 Old Claims Process at Insurance Looks good, cut them a check Headquarters Mainframe 31 Old Claims Process at Insurance Customer has time to exaggerate or make up claims Supervisor doesn’t assign cases based on expertise Headquarters Mainframe 32 New Claims Process at Insurance Customer must phone in claim Headquarters Mainframe Claim handler prints and sends to customer for 33 signature and also saves electronic copy of claim New Claims Process at Insurance Signed customer document scanned and stored on image server Headquarters Mainframe Mainframe program assigns complexity and expertise rating from 0 to 20 34 New Claims Process at Insurance Headquarters Mainframe Each claims handler has an expert rating to determine what claims they can work on 35 Manufacturing • Poor customer service, poor manufacturing coordination due to a lack of shared information about each of the plant’s capabilities, capacities, and schedules. • BPR: Centralize decision-making, software selection drove client/server • Implemented a global client/server application for customerordering, manufacturing planning, plant scheduling, and product specification. • Improved customer service from 3 weeks to 4 hours, decrease manufacturing cycle from 27 to 19 days. • Huge success, although senior management credit manufacturing with the improvements 36 Manufacturing 37 Manufacturing Can I order 60,000 wafers with these 200 parameters? 38 Manufacturing Let me check, I’ll get back to you… 39 Manufacturing Can you do 60,000 wafers? 40 Manufacturing Well, let me see if I can get some crystal... 41 Manufacturing Hey Joe--can I have bunch of crystal by next week? 42 Manufacturing I can give you half what you need…try Ben 43 Manufacturing Ben--do you have x amount of crystal to ship by next week? 44 Manufacturing No, but I can make some within 3 weeks. Want it? 45 Manufacturing Let me get back to you 46 Manufacturing Susan, can I have x amount of crystal? 47 Manufacturing Sure I can ship today! 48 Manufacturing I can make 30,000 wafers in 3 months, but the rest will take 5 months. 49 Manufacturing Let me try some other plants first... 50 Manufacturing THREE WEEKS LATER…. Great news we can do it! 51 Manufacturing Thanks but I went with your competitor--he gave me a quote the next day. 52 Manufacturing This is what you are going to make and when! 53 Manufacturing Yipee!!! Sir I am happy to confirm your order within 4 hours! 54 Baking • $8 million per year on mainframe outsourcing contract perceived as too expensive. • Migration of mainframe transaction processing systems to client/server for 35 bakeries to client/server to reduce costs. • No changes in organizational structure or processes. • Pilot project a technical success • • Project canceled when company was sold New management believes in “letting each bakery select its own technology” 55 Baking Headquarter Mainframe Daily sales, returns, orders 56 Baking Headquarter Mainframe Daily baking schedules, driver instructions, prices, discounts, etc. 57 Baking Headquarter Mainframe Daily baking schedules, driver instructions, prices, discounts, etc. 58 Original Plan at Baking 59 Denver Pilot at Baking Big success! 60 Revised Plan at Baking 61 PROJECT OUTCOMES All three projects were late and over-budget: Company Expected Cost Manufacturing Baking Insurance $ 6 million $ 7 million £4.5 million Actual Cost Number of Months Late $ 12 million 24 Project Canceled Project Canceled £6 million + 10 But are they successes or failures? 62 PROJECT JUSTIFICATION PRACTICES Not just cost/benefit analysis Contribution to business strategy: develop new distribution channels mass customization Process improvement: quicker manufacturing time faster product development improved customer service reduced inventory levels Opportunity costs: what happens if we don’t do this project? 63 REASONS FOR COST/TIME OVERRUNS Companies had to develop (or help the vendor develop) the client/server technology. TECHNICAL UNCERTAINTY: Unexpected incompatibilities among vendor products Debug vendor’s new client/server applications Technical cracks in back-ups, roll-outs, and check-points REQUIREMENT UNCERTAINTY: UNREALISTIC EXPECTATIONS OF PROJECT DELIVERY UNREALISTIC COST/BENEFIT ANALYSIS IN ACQUISTION REQUEST 64 PROJECT JUSTIFICATION ? Company Manufacturing Baking Insurance Only 1 of the companies quantified the benefits Quantifiable Benefits $500,000 a year savings to terminate mainframe outsourcing $3 million a year savings to terminate mainframe outsourcing (however, the outsourcing vendor reduced bill by this amount during c/s development) Financial Guesses Increase sales by $15 million per year 2% reduction in costs of claims (which translates to about £5 million annually). 65 Lesson: Companies needed to secure top management support in the project, primarily by having a senior-level project sponsor. TYPE OF SUPPORT Project Champion Project Sponsor SILICON X BAKING X INSURANCE X X X 66 Lesson: Companies needed to secure top management support in the project, primarily by having a senior-level project sponsor. • Project Champion: “Champions are managers who actively and vigorously promote their personal vision for using IT, pushing the project over or around approval and implementation hurdles. They often risk their reputations in order to ensure the innovation’s success” (Beath, 1991, p. 355) • Project Sponsor: “Sponsors have the funds and authority to accomplish their goals” (Beath, 1991, p. 355) (A less active and less enthusiastic role compared to a champion) 67 Lesson: Business strategy drove technology selection in 2 companies Company Manufacturing Business Process Re-engineering Baking none: “It’s a business problem that we solved using this technology as opposed to the other way around. [The MIS Director] thinks, from what he has seen, if you are installing C/S just to install C/S, you are crazy, because there is a high learning curve to get over.” --Strategic Manager of Marketing “You can buy a $100,000 processor that if you want to use MIPS or throughput, so you can get a pretty sophisticated piece of equipment for $100,000. To get the equivalent in IBM, older generation, you are talking $700,000 to $1,000,000. So what I presented to management was, ‘Here we are on this path with this investment base and its dropping at this rate. Here’s a new technology base, a lower entry fee, but its cost is dropping. How do we get from this curve to that curve without killing ourselves and without making a zillion dollar investment?’”-- Head of IT Department Insurance “We are re-engineering the business process. It takes a large number of business functions, so for example, process work flow handling. It has business rules in it for processing claims. You take it immediately into a PC arena. Also because we are processing rules, it takes us into the client/server arena. We are actually holding some data locally to assign work flow.”-- Manager of the Business Technology Department 68 Lesson: All companies hired some outside experts, to varying success Company Manufacturing Baking Insurance Supplier Relationship Received free vendor resources in exchange for helping to debug new client/server system. Knowledge about the new technology was transferred so that internal IT staff were in a position to support it. Baking hired a consulting firm to help with implementation: “I got the feeling they were only one chapter ahead of us in the book.” At first, Insurance relied on vendor salespeople. After technical incompatibilities arose, they hired an outsourcing consulting company to review/verify technical design. 69 Lesson: All companies hired some outside experts, to varying success Lesson: From prior research on 116 sourcing case studies, the development of immature technologies were the worst to outsource because specifications were uncertain and no transfer of knowledge occurred. The best sourcing model for immature technologies: buy-in vendor resources 70 Lesson: Each company implemented incrementally Company Implementation Strategy Manufacturing Implemented 3 phases, each with added-value to business Phase I: central specifications, scheduling Phase II: 700 management reports Phase III: Front-end customer orders Baking Pilot at one bakery Insurance Mock-up “model” office Benefits of incremental implementation: Verify user requirements Reduce risk Recover from mistakes early Demonstrate value to secure management approval for full-scale implementation Gain user acceptance of the system 71 QUESTIONS TO THINK ABOUT: Should success be based “on time” “on budget” “to functionality?” • Baking is a Type III project that met all objectives • Insurance is a Type II, mediocre results • Manufacturing is a Type II, excellent results Which projects “Did the thing right?” Which projects “Did the right thing?” 72 Mastering the Next Wave of Enterprise Systems: Leveraging Lessons From ERP Top Management is engaged in the project, not just involved Project leaders are veterans with proven track records and team makers are best in the business empowered to make decisions Buy-in vendor resources to fill gaps in expertise and transfer their knowledge Change management goes hand in hand with project management A satisficing mindset prevails. By Carol Brown and Iris Vessey, MISQE, 2003. 73 Three “Successful” ERP Implementations “Material” “Valvo” “Asea” Transform a small branch to profit center with own selling & distribution Revenues $3.5 Billion $430 Million Reason for ERP Replace 200 legacy systems to provide global business processes Replace lousy legacy systems running on old mainframe packages Cost of ERP $100 million $17 million $150,000 Time Line April 1995 Late 1997 Sept 1996 Year end 1997 Sept 2000 Year end 2000 Time ~2.75 years; Late & over budget ~1.5 years ~4 months Sooner & on budget 50% increase in inventory turns 20% reduction in admin costs $millions in logistical costs Y2k compliance 35% reduction in inventory costs Better order fulfillment rates Outcomes On time & under budget (but increased 30% after planning phase) Allowed new profit center to get up and running with selling & inventory 74 Noted Practices “Material” “Valvo” “Asea” Adoption Diffusion Lesson Early adopter achieves Success by learning by Doing with “freedom” To fail culture Early majority achieves success by benchmarking successful early adopters Late majority achieve success by applying templates. Buy-in of vendor resources Big 5 firm with 50 consultants on project Convinced consulting firm to help them go big bang Used shared services experts located in Malaysia with R/3 expertise Change Management Lead by organizational development & training specialists 900 power users Cash bonus Stock options Project co-lead along with IT and business; Documented new work, newly automated work, eliminated work; work transferred to another group Bonus pay to everyone Business co-led with IT was also responsible for change management 75 Session Objectives Business solutions drive technology selection Secure top management active support Select high powered project manager with proven Ensure to track record Value for IT Involve & incent knowledgeable users Spend Don't judge success solely based on time to budget Buy-in outside expertise to transfer learning Implement incrementally Manage scope creep, perhaps using 80/20 satisficing rule Beware of Mythical Man Month Doing The Thing Right: Robust Best Practices for Project Management 76